GPT4o-mini -> No images supported?

API
1

Hello everyone!

I built a scenario in Make that is supposed to extract information from supplier invoices (jpeg files).

4o-mini seemed perfect for that. I upload the file to OpenAI and then provide the assistant the file id:

image
image884×886 53.3 KB

But I can’t get it to work. I always get an output along the lines of ‘Please provide a file / Can’t access the file’.

With regular GPT4o it works though.

Aren’t they both supposed to support image files? Where’s the error on my part?

Thanks a lot in advance!

2

Welcome to the dev forum @s_tom

Images can only be supplied in messages as urls to the images on the web or File IDs of the images uploaded via the File API with purpose="vision".

Here’s docs for further details.

3 Likes
3

Hi, thanks for your response! I now changed the purpose to ‘Vision’ but the error still persists. Do you have any idea why that could be? Thank you!